Terra: Imperative-Symbolic Co-Execution of Imperative Deep Learning Programs

TL;DR

Terra is a novel system that enables imperative deep learning programs to be executed with symbolic graph optimization, supporting all Python features and outperforming existing systems in speed and coverage.

Contribution

Terra introduces a co-execution approach that decouples Python features from DNN operations, allowing full imperative support with symbolic performance optimization.

Findings

Terra speeds up all tested imperative DL programs.

AutoGraph fails on five out of ten programs.

Terra achieves broader execution coverage.

Abstract

Imperative programming allows users to implement their deep neural networks (DNNs) easily and has become an essential part of recent deep learning (DL) frameworks. Recently, several systems have been proposed to combine the usability of imperative programming with the optimized performance of symbolic graph execution. Such systems convert imperative Python DL programs to optimized symbolic graphs and execute them. However, they cannot fully support the usability of imperative programming. For example, if an imperative DL program contains a Python feature with no corresponding symbolic representation (e.g., third-party library calls or unsupported dynamic control flows) they fail to execute the program. To overcome this limitation, we propose Terra, an imperative-symbolic co-execution system that can handle any imperative DL programs while achieving the optimized performance of symbolic graph execution. To achieve this, Terra builds a symbolic graph by decoupling DL operations from Python features. Then, Terra conducts the imperative execution to support all Python features, while delegating the decoupled operations to the symbolic execution. We evaluated the performance improvement and coverage of Terra with ten imperative DL programs for several DNN architectures. The results show that Terra can speed up the execution of all ten imperative DL programs, whereas AutoGraph, one of the state-of-the-art systems, fails to execute five of them.